IEEM 230. Industrial Data Systems A Database is a collection of information [Data, Procedural/behavioral knowledge] about a system.

A computerized Database Management System (DBMS) is a collection of programs that help in: Organizing the information, Defining the database, Constructing the database, and Manipulating its contents.

Why do we need to use a DBMS (what are the advantages)?

1. Advantages arising out of use of a computer Easy storage of large amounts of data; Better protection against data destruction; Faster Access of information.

2. Restricting Unauthorized Access Example: A Database in a bank

3. Controlling Redundancy Redundant information: same information being unnecessarily repeated in more than one storage locations.

Example: HKUST databases, if ARR and IEEM dept maintain separate data: ARR records need to store: Dept records need to store:

What is the problem with redundant data? - Needs extra storage space. - Extra effort in creating & maintenance of information. - Impossible to maintain data consistency.

Controlled redundancy is, however, necessary, and useful! Why? - It may help to recover lost data in case one data storage is destroyed. - It can allow faster access times for some information.

4. Persistent Storage of Program Objects (and Data Structures) In many programming applications, very large and complex data structures are used to generate information, which is used to perform some reasoning. Once the program terminates, all these data structures are lost (removed from the computer memory), unless they are explicitly stored in flat files. Modern, Object Oriented Databases allow the direct storage of such data structures in the DB, for easy and quick retrieval for re-use.

Example: A factory scheduling software.

5. Providing multiple user interfaces

6. Enforcing Data Integrity Constraints

This can be done a several levels: data type, correct range of values for some data, constraints between two different data entries, etc.

Relational Database Systems: Architecture and Concepts

There are several types of DBMSs: Relational DB, Object Oriented DB, Hierarchical DB, Distributed DB, etc. In this course, we will study only Relational DBMS, which is the most commonly used DBMS in modern industry.

The structure of a database is described by the use of data models.

Data Model contains: -A description of the data types, their relationships and constraints. - Set of basic operations used to manipulate the data in the DB. - The behavior, i.e., a set of user-defined operations allowed on the data

There are different methods to model data; we will study one method: Entity-Relation Model (ER Model) in more detail later on.

Who uses a DBMS ? Almost everyone ! Examples: Sales clerk at Park-and-Shop Design engineer in a company Accountant, Bank clerk Clerk in a personnel office IT officer in a company Web-page designer The following figure shows different types of users of a DBMS, and how they use the DB.

Types of Users DBA Staff: People who are involved with the design, implementation, and maintenance of the DBMS.

Parametric end users: People who use an application program interface to get data from a DB (e.g. when you check your bank account balance from the ATM machine).

Casual end users: People with direct access to (some tables) in a database, and who can get this data by writing and executing queries.

Applications programmers: IT engineers who will write interface programs that will access some data in a DB. Such programs can allow parametric users to do data entry, or to access some information.

Brief explanation of the terms used in the figure

The Stored Data Manager: This module controls the access to all data in a DBMS. This includes all data in the DB, as well as all data in the catalogs.

Catalog (data dictionary): The catalog stores information about the data stored in a database. For example, a DB may have a table called Student that stores information with fields called Name, DOB, Phone, Address, Dept. The data dictionary could store information such as: Name is a stored as a character string, with last name followed by first names; DOB is date of birth, in the format DD-MM-YYYY; etc.

Privileged Commands: Special authority is required to execute these commands, e.g. if you want to delete the entire DB.
5

DDL Statements, DDL Compiler: DDL: Data definition Language. The description of the DB, namely the Data Dictionary, must be done by using a set of rules (e.g. each table must have a description which is text; each column of the table must have a name, and must define the format, etc). To force these rules, all DB definitions must be made in a language that can be processed by a computer. This is called the Data Definition Language.

The Run time database processor: This is the main function of the DBMS which executes operations such as: storage of data, retrieval of data, etc.

Query Compiler: DB users can access the DB by using a high-level language (which can be typed using a keyboard); these access commands are called queries. The query compiler converts each query into machine language that can be processed by the computer.

Application Programming: Application programmers are users who write computer programs that are linked to the DB. Application programs can be written using any computer language (e.g. C++, VisualBasic, PHP, etc.) Such programs must connect to a DBMS, in order to manipulate the data stored in the DB. Thus, a C++ application program will have normal C++ statements, and also some statements in a language (called DML, or data manipulation language) that can be understood and processed by the DBMS. A pre-compiler separates the C++ and the DML statements in the applications program. The C++ statements are compiled and executed by the computer processor; the DML statements are sent to the DBMS, which will then process them, and send the output data back to the application program.